Recognition of virus-infected cells by natural killer cell clones is controlled by polymorphic target cell elements.

Natural killer (NK) cells provide a first line of defense against viral infections. The mechanisms by which NK cells recognize and eliminate infected cells are still largely unknown. To test whether target cell elements contribute to NK cell recognition of virus-infected cells, human NK cells were cloned from two unrelated donors and assayed for their ability to kill normal autologous or allogeneic cells before and after infection by human herpesvirus 6 (HHV-6), a T-lymphotropic herpesvirus. Of 132 NK clones isolated from donor 1, all displayed strong cytolytic activity against the NK-sensitive cell line K562, none killed uninfected autologous T cells, and 65 (49%) killed autologous T cells infected with HHV-6. A panel of representative NK clones from donors 1 and 2 was tested on targets obtained from four donors. A wide heterogeneity was observed in the specificity of lysis of infected target cells among the NK clones. Some clones killed none, some killed only one, and others killed more than one of the different HHV-6-infected target cells. Killing of infected targets was not due to complete absence of class I molecules because class I surface levels were only partially affected by HHV-6 infection. Thus, target cell recognition is not controlled by the effector NK cell alone, but also by polymorphic elements on the target cell that restrict NK cell recognition. Furthermore, NK clones from different donors display a variable range of specificities in their recognition of infected target cells.

Diversity and structure of human T cell receptor delta chain genes in peripheral blood gamma/delta-bearing T lymphocytes.

We have investigated the diversity and repertoire of human TCR delta chain variable gene segments in the human peripheral blood CD4- CD8- (double-negative) population, using rearrangement and expression studies and sequence analyses. 20 TCR delta DNA clones were derived from the RNA of bulk-cultured double-negative T cells and their nucleotide sequences determined. These clones can be classified into six different V delta subfamilies. The distribution, however, was uneven in these cells, with 16 of 20 being derived from the V delta 1 (9) and V delta 2 (7) subfamilies. The remaining subfamilies, V delta 3, V delta 4, V delta 5, and V delta 6, were only represented by one clone each. The majority of these subfamilies seem to consist of a single member, in contrast with the closely linked V alpha subfamilies, which, in most cases, consist of multiple members. Our findings suggest that only a limited number of V delta genes are used in human peripheral blood double-negative T cells and that two major V delta subfamilies (V delta 1 and V delta 2) are used more frequently. Sequence comparison of our cDNA clones to V alpha clones indicates that there is no overlap in usage of V alpha and V delta gene segments, except for the V delta 4 (V alpha 6) subfamily. Comparison of the different V delta sequences suggests that the majority of the sequence diversity is concentrated in the junctions between V, D, and J segments and results from extensive N region diversity.

Regulation of hematopoiesis in vitro by alloreactive natural killer cell clones.

Natural killer (NK) cells lyse autologous and allogeneic target cells even in the absence of major histocompatibility complex (MHC) class I antigens on the target cells. Recently, however, human allospecific NK cell clones have been generated that recognize at least five distinct specificities inherited recessively and controlled by genes linked to the MHC. Because the genetic specificity of these alloreactive NK cells in vitro appears analogous to that of in vivo NK cell-mediated murine hybrid resistance, i.e., the rejection of parental bone marrow in irradiated F1 animals, we tested the ability of human alloreactive NK clones to recognize allogeneic hematopoietic progenitor cells. NK cells from two specificity 1 alloreactive NK clones, ES9 and ES10, significantly and often completely suppressed colony formation by purified peripheral blood hematopoietic progenitor cells from specificity 1-susceptible donors, but had no significant effect on the cells of specificity 1-resistant donors. Activated polyclonal NK cells were less efficient than the NK clones in inhibiting colony formation and had a similar effect on cells from both specificity 1-susceptible and -resistant donors. The alloreactive NK clones produced cytokines with a suppressive effect on in vitro hematopoiesis, such as interferon gamma (IFN-gamma) and tumor necrosis factor alpha (TNF-alpha), when exposed to phytohemagglutinin blasts from specificity 1-susceptible, but not -resistant donors. However, the mechanism by which alloreactive NK cells inhibit colony formation is more consistent with a direct cytotoxic effect than with the production of inhibitory cytokines because antibodies (anti-IFN-gamma, alpha-TNF-alpha, and -lymphotoxin) that completely blocked the inhibition by polyclonal NK cells had only a minimal effect on the inhibition by the alloreactive clones. Moreover, the alloreactive clones were directly cytolytic in a 51Cr release assay against enriched preparations of peripheral blood progenitor cells from specificity 1-susceptible donors. These data indicate that the alloreactive NK cells are likely the human counterpart of the cells mediating murine hybrid resistance and that these cells might play clinically important roles in rejection or in graft-versus-leukemia reactions after allogeneic bone marrow transplantation.

CD69-mediated pathway of lymphocyte activation: anti-CD69 monoclonal antibodies trigger the cytolytic activity of different lymphoid effector cells with the exception of cytolytic T lymphocytes expressing T cell receptor alpha/beta.

The effect of anti-CD69 monoclonal antibodies (mAbs) on the induction of the cytolytic activity in different types of lymphoid effector cells has been investigated. Three anti-CD69 mAbs, including the reference mAb MLR3 and two new mAbs (c227 and 31C4), have been used. All cloned CD3-CD16+ natural killer (NK) cells belonging to different subsets (as defined by the surface expression of GL183 and/or EB6 antigens) were efficiently triggered by anti-CD69 mAbs and lysed P815 mastocytoma cells in a redirected killing assay. Triggering of the cytolytic activity could also be induced in CD3-CD16- NK clones, which fail to respond to other stimuli (including anti-CD16, anti-CD2 mAbs, or phytohemagglutinin). A similar triggering effect was detected in T cell receptor (TCR) gamma/delta+ clones belonging to different subsets. On the other hand, anti-CD69 mAbs could not induce triggering of the cytolytic activity in TCR alpha/beta+ cytolytic clones. Since all thymocytes are known to express CD69 antigen after cell activation, we analyzed a series of phenotypically different cytolytic thymocyte populations and clones for their responsiveness to anti-CD69 mAb in a redirected killing assay. Again, anti-CD69 mAb triggered TCR gamma/delta+ but not TCR alpha/beta+ thymocytes. Anti-CD69 mAb efficiently triggered the cytolytic activity of "early" thymocytes lines or clones (CD3-4-8-7+), which lack all other known pathways of cell activation. Thus, it appears that CD69 molecules may initiate a pathway of activation of cytolytic functions common to a number of activated effector lymphocytes with the remarkable exception of TCR alpha/beta+ cytolytic cells.

Identification of four subsets of human CD3-CD16+ natural killer (NK) cells by the expression of clonally distributed functional surface molecules: correlation between subset assignment of NK clones and ability to mediate specific alloantigen recognition.

In previous studies we identified a surface molecule (termed GL183) capable of mediating cell activation and selectively expressed by a subset of human CD3-CD16+ natural killer (NK) cells. In this study we analyzed whether other subset-specific functional molecules were expressed in GL183- NK cells. To this end, mice were immunized with the PE29 (CD3-CD16+GL183-) NK clone. Monoclonal antibodies (mAbs) were selected by screening the hybridoma supernatants for their ability to trigger the cytolytic activity of clone PE29 against the human myelomonocytic leukemia U937. The EB6 mAb (IgG1) triggered the PE29 clone, but not a GL183+ clone used as a control. EB6+ cells ranged between 1 and 13% of peripheral blood lymphocytes and were largely included in the CD3-CD16+CD56+ cell populations (only less than 2% of EB6+ cells were CD3+). Analysis of resting or activated CD3-CD16+ populations, or clones for the expression of EB6 or GL183 mAbs, allowed us to identify four distinct, phenotypically stable, NK subsets (EB6+GL183-; EB6+GL183+; EB6-GL183+; EB6-GL183-). Similar to GL183 mAb, the EB6 mAb selectively triggered the NK subset expressing the corresponding surface antigen to lyse human tumor cell lines including U937, IGROV-I, M14, and A549. In addition, EB6 mAb sharply inhibited the cytolytic activity of EB6+ clones against P815, M12, and P3U1 murine target cells. In EB6+GL183+ ("double-positive") clones both EB6 and GL183 mAb inhibited the redirected killing of P815 cells induced by anti-CD16, anti-CD2 mAbs and phytohemagglutinin (PHA). Similar to GL183 molecules, molecules precipitated by EB6 mAb were represented by either single 58-kD chain or double chains of 55 and 58 kD (with no detectable differences in EB6+GL183- or EB6+GL183+ clones). In sequential immunoprecipitation experiments using the double-positive clones CEG52 and CA25.50, preclearing of cell lysates with EB6 or GL183 mAb removed only EB6 or GL183 molecules, respectively, thus indicating that the two antigenic determinants are carried by two distinct molecules. Peptide map analysis indicated that EB6 (or GL183) molecules precipitated from double-positive clones were identical to the corresponding molecules isolated from single-positive ones. On the other hand, comparison of the EB6 and GL183 maps revealed peptides that were unique to each molecule, although most of the major peptides migrated to identical positions. We further investigated whether correlation existed between the phenotypic assignment of NK clones and their ability to mediate specific lysis of normal allogeneic cells.(ABSTRACT TRUNCATED AT 400 WORDS)

Antigen recognition by human T cell receptor gamma-positive lymphocytes. Specific lysis of allogeneic cells after activation in mixed lymphocyte culture.

These experiments were designed to define the ability of human TCR-gamma+ cells to recognize allogeneic cells. TCR-gamma+-enriched populations were obtained by treating peripheral blood E-rosetting cells with anti-CD4 and anti-CD8 mAbs. The resulting populations were CD2+4-8- expressed variable proportions of CD3+ cells (40-90%), and did not react with the WT31 mAb, which is specific for a framework determinant of the alpha/beta heterodimer that serves as receptor for antigen on most human T lymphocytes. After mixed lymphocyte culture with irradiated allogeneic cells for 7 d and 3 additional days in rIL-2 (100 U/ml), cells underwent proliferation in three of five individuals tested. In addition, MLC-derived cells lysed 51Cr-labeled PHA-induced blasts derived from the allogeneic cells used as stimulator, but not allogeneic unrelated or autologous blast cells. No cytotoxicity against autologous or allogeneic target cells could be induced by culturing CD3+4-8-WT31- lymphocytes in MLC with irradiated autologous cells. Surface iodination of allogeneic MLC-activated CD3+4-8-WT31- cells followed by lysis in 1% digitonin and immunoprecipitation with anti-CD3 mAb indicated that the CD3-associated molecules consisted of a major 45-kD band and a minor band of 43 kD. Northern blot analysis showed that mRNA for the gamma chain was expressed at high levels, whereas mRNAs for alpha and beta chains were missing. These data support the notion that TCR-gamma rather than TCR-alpha/beta is expressed in allospecific CD3-4-8-WT31- cell populations. Clones were further derived from MLC-stimulated CD3+4-8-WT31- populations. All the seven clones studied in detail maintained the surface phenotype as well as the cytolytic pattern of the original MLC populations, thus only specific allogeneic PHA-induced blasts were lysed. NK-sensitive as well as NK-resistant tumor targets were variably susceptible to lysis; therefore, specific cytolytic activity against allogeneic cells was not necessarily linked to the expression of MHC-nonrestricted cytotoxicity against tumor cells.

A monoclonal antibody specific for a common determinant of the human T cell receptor gamma/delta directly activates CD3+WT31- lymphocytes to express their functional program(s).

In an attempt to select mAbs specific for human TCR-gamma/delta, a polyclonal CD3+ 4-8-WT31- (TCR-gamma/delta+) cell line (MV1) was used for mice immunization. An mAb, termed BB3, reacted with MV1 cells but not with a large panel of CD3+ WT31+ (TCR-alpha/beta+) cell populations or clones. In addition, BB3 mAb reacted with the majority of CD3+ WT31- clones derived from six different donors. Double-color fluorescence experiments and FACS analysis showed that BB3+ cells were restricted to the CD3+ fraction of peripheral blood lymphocytes; in addition, in several donors the percentages (0.5-8% of total PBL) of BB3+ cells paralleled those of CD3+ WT31- cells. Surface molecules recognized by BB3 were susceptible to antibody-induced modulation; in addition, cell treatment with either BB3 or anti-CD3 mAb caused the simultaneous downregulation of the two molecules. That BB3 molecules are physically linked to CD3 antigen was further supported by immunoprecipitation experiments. Thus, under conditions that preserve the TCR-CD3 association, both BB3 and anti-CD3 mAb precipitated from 125I-labeled MV1 cells the same set of molecules. These consisted in the 18-28-kD CD3 molecules and in three bands of approximately 44, 42, and 38 kD under reducing conditions. When cell lysis was performed in 1% NP-40, the molecules immunoprecipitated by BB3 mAb were represented by an 80-kD band under nonreducing conditions, which resolved, under reducing conditions, in the three 44-, 42-, and 38-kD bands. Similar disulphide-linked forms of the TCR molecules were revealed in all of the other eight CD3+ WT31- BB3+ clones analyzed. Analysis of TCR molecules by electrophoresis (NEPHGE) showed that BB3 or anti-CD3 precipitated a 44-kD molecule displaying a basic PI (approximately 7.5) and two more acidic proteins (PI approximately 6) with a mol mass of 42 and 38 kD. Studies aimed to define whether stimuli directly acting on TCR-gamma/delta could induce CD3+ WT31- cell activation revealed that (a) In the presence of PMA, soluble BB3 mAb induced IL-2 production by MV1 cell line and by three other CD3+ WT31- BB3+ clones analyzed. (b) BB3 mAb-producing hybridoma used as triggering target, was efficiently lysed by CD3+ WT31- BB3+ effector cells (but not by CD3+ WT31+ BB3- conventional CTL). (c) Soluble BB3 mAb induced CD3+ WT31- BB3+ effector cells to lyse the Fc receptor-positive P815 target cells. (d) BB3-TCR-gamma/delta interaction on CD3+ WT31- BB3+ cells induced a rapid increase of [Ca2+]i levels, similar to that observed in response to anti-CD3 mAbs.

Human peripheral blood lymphocytes bearing T cell receptor gamma/delta. Expression of CD8 differentiation antigen correlates with the expression of the 55-kD, C gamma 2-encoded gamma chain.

We analyzed the CD3-associated molecules present on peripheral blood-derived TCR-gamma/delta+ clones that express CD8 surface antigens. Clones were derived by limiting dilution from CD3+WT31- FACS-purified populations derived from several donors. Eight of greater than 300 TCR-gamma/delta+ clones analyzed expressed CD8 and reacted with delta-TCS-1 mAb. Cell numbers suitable for more detailed analyses could be obtained from four clones, including one derived from thymus. Analysis of CD3-associated TCR molecules immunoprecipitated by anti-Leu-4 (anti-CD3) mAb under conditions that preserve the CD3/TCR association (1% digitonin) showed a predominant 55-60-kD molecule both under reducing and nonreducing conditions. On the other hand, the delta-TCS-1-reactive molecules immunoprecipitated from 25 CD3+ delta-TCS-1+ CD8- clones, in all instances, displayed a 40-44-kD mol mass. In two-dimensional PAGE, TCR-gamma molecules precipitated from delta-TCS-1+ CD8+ clones appeared more acidic than those of BB3+ or delta-TCS-1+ CD8+ clones. Southern analysis confirmed that this type of non-disulphide-linked TCR-gamma/delta is also coded for by the C gamma 2 gene segment.

P58 molecules as putative receptors for major histocompatibility complex (MHC) class I molecules in human natural killer (NK) cells. Anti-p58 antibodies reconstitute lysis of MHC class I-protected cells in NK clones displaying different specificities.

Human CD3-16+56+ natural killer (NK) cells have been shown to display a clonally distributed ability to recognize major histocompatibility complex (MHC) class I alleles. Opposite to T lymphocytes, in NK cells, specific recognition of MHC class I molecules appears to induce inhibition of cytolytic activity and, thus, to protect target cells. Since a precise correlation has been established between the expression of the NK-specific GL183 and EB6 surface molecules (belonging to the novel p58 molecular family) and the specificity of NK clones, we analyzed whether p58 molecules could function as receptors for MHC in human NK cells. NK clones displaying the previously defined "specificity 2" and characterized by the GL183+EB6+ phenotype, specifically recognize the Cw3 allele and thus fail to lyse the Fc gamma R+ P815 target cells transfected with Cw3. On the other hand, NK clones displaying "specificity 1" and expressing the GL183-EB6+ phenotype failed to lyse Cw4+ target cells. Addition of the F(ab')2 fragments of either GL183 or EB6 mAb as well as the XA141 mAb of IgM isotype (specific for the EB6 molecules) completely restored the lysis of Cw3-transfected P815 cells by the Cw3-specific NK clones EX2 and EX4. Similarly, both the entire EB6 mAb, its F(ab')2 fragment and the XA141 mAb reconstituted the lysis of C1R, a Fc gamma R- target cell expressing Cw4 as the only serologically detected class I antigen. Thus, it appears that masking of different members of p58 molecules prevents recognition of "protective" MHC class I alleles and thus the delivering of inhibitory signals. Further support to the concept that p58 molecules represent a NK receptor delivering a negative signal was provided by experiments in which the entire anti-p58 mAbs (of IgG isotype) could inhibit the lysis of unprotected Fc gamma R+ P815 target cells, thus mimicking the inhibitory effect of MHC class I molecules.

A novel surface antigen expressed by a subset of human CD3- CD16+ natural killer cells. Role in cell activation and regulation of cytolytic function.

The GL183 mAb was obtained by immunizing BALB/c mice with the E57 clone (CD7+CD2+CD3-CD16+CD56+) derived from human peripheral blood NK cells. In human peripheral blood, GL183-reactive cells ranged between 2 and 12% (mean 6.5%) in 10 different donors. Double fluorescence and FACS analysis showed that GL183+ cells were consistently included in the CD56+ or CD16+ cell populations. Moreover, since only a fraction of CD56+ or CD16+ cells (approximately 40%) coexpressed GL183 surface antigen, reactivity with GL183 mAb appears to define two subsets within the CD3- lymphocyte population expressing NK cell markers. Although, the majority of GL183+ cells were CD3-, approximately 1% expressed CD3 surface antigens. As shown by clonal analysis, these infrequent CD3+GL183+ cells coexpressed CD56 and CD16 antigens. Cloning of CD3-GL183+ or CD3-GL183- cell populations under limiting dilution conditions yielded clonal progenies that maintained their original surface phenotype. Therefore, expression or lack of expression of GL183 surface antigens represents a stable phenotypic property of a subset of human CD3- NK cells. Immunoprecipitation experiments and two-dimensional PAGE analysis indicated that GL183-reactive molecules were represented in different clones either by a single 58-kD chain or, more frequently, by two chains of approximately 55 and approximately 58 kD, respectively. Analysis of GL183+ or GL183- NK clones for their ability to lyse human (IGROV I) or murine (P815) tumor target cells indicated that GL183- clones were, on average, fivefold more efficient in inducing target cell lysis. GL183+ and GL183- clones produced comparable levels of TNF-alpha in response to PHA plus PMA or anti-CD16 mAb plus PMA. Importantly, production of TNF-alpha was also induced by stimulation of GL183+ clones with GL183 mAb plus PMA. These data indicated that GL183 antigen could mediate cell triggering. This concept was confirmed by the analysis of Ca2+ mobilization, as GL183 mAb induced (in GL183+ clones) increments of [Ca2+]i comparable with those induced by PHA. Moreover, GL183 mAb, or its F(ab')2 fragments, strongly enhanced the cytolytic activity of GL183+ clones against a panel of human tumor target cells, including U937, Raji, IGROV I, M14, and A549. In contrast, GL183 mAb, but not the F(ab')2 fragments, sharply inhibited the cytolytic activity of the same clones against P815, M12, and P3U1 murine target cells. In this case, the effect of GL183 mAb (inhibition) was opposite that of PHA or of stimulatory anti-CD2 or anti-CD16 mAbs, which consistently enhanced the target cell lysis.(ABSTRACT TRUNCATED AT 400 WORDS)

Specific recognition of human CD3-CD16+ natural killer cells requires the expression of an autosomic recessive gene on target cells.

We analyzed the recently defined ability of CD3-CD16+ cells to specifically recognize and lyse normal allogeneic target cells (PHA-induced blasts). The susceptibility to lysis by a given alloreactive natural killer (NK) clone ("1 anti-A") was expressed by PHA blasts derived from 9 of 38 random donors analyzed. In all instances, the specific lysis of "susceptible" target cells was greater than 35% while that of "nonsusceptible" targets was less than 6% at an E/T cell ratio of 5:1. In addition to 1 anti-A, A anti-1 specific CD3-CD16+ clones could also be isolated from the reverse MLC combination. The relationship existing between lysis of normal allogeneic cells or tumor cells by the same CD3-CD16+ effector cell has been investigated: 1 anti-A specific CD3-CD16+ clones lysed PHA blasts of three of six cancer patients, while they lysed fresh tumor cells (ovarian carcinoma) from all six patients. The type of inheritance of the character "susceptibility to lysis" was analyzed in representative families. This analysis revealed that the character is inherited in an autosomic recessive fashion, and it is therefore different from MHC. We further investigated the type of segregation of the opposite character "resistance to lysis" (which is inherited in a dominant mode). The finding that this character segregated in all donors expressing given MHC haplotypes indicated that the gene regulating the expression of the NK-defined alloantigen is present on chromosome 6.

In vitro proliferation and cloning of CD3- CD16+ cells from human thymocyte precursors.

Purified CD3-4- thymocytes were obtained by depletion of CD3+ and CD4+ cells from fresh thymocyte suspensions. 5-15% of these cells were found to express CD16 antigen, while other natural killer (NK) cell markers were virtually absent. Double fluorescence analysis revealed that 20-40% of thymic CD16+ cells coexpressed CD1, while approximately half were cyCD3+. When cultured in the presence of peripheral blood lymphocytes and H9 leukemia cell line as a source of irradiated feeder cells and interleukin 2 (IL-2), CD3-4- thymocytes underwent extensive proliferation. In addition, after 1-2 wk of culture, 30-50% of these cells were found to express CD16 surface antigen. Cloning under limiting dilution conditions of either CD3-4- or CD3-4-16- thymocytes in the presence of irradiated H9 cells resulted in large proportions (approximately 50%) of CD16+ clones. On the basis of the expression of surface CD16 and/or cyCD3 antigen, clones could be grouped in the following subsets: CD16+ cyCD3+; CD16+ cyCD3-; CD16- cyCD3+; and CD16- cyCD3-. All clones expressed CD56 surface antigen, displayed a strong cytolytic activity against NK sensitive (K562) and NK-resistant (M14) target cells, and produced IFN-gamma and tumor necrosis factor, but not IL-2. Similar to peripheral NK cells, thymic CD16+ cells expressed transcripts for CD16 and for CD3 epsilon (Biassoni, R., S. Ferrini, I. Prigione, A. Moretta, and E.O. Long, 1988. J. Immunol. 140:1685.) and zeta chains (Anderson, P., M. Caligiuri, J. Ritz, and S.F. Schlossman. 1989. Nature [Lond.]. 341:159). Therefore, it appears that cells that are phenotypically and functionally similar to CD3- CD16+ NK cells may arise from immature thymocytes.

Specific lysis of allogeneic cells after activation of CD3- lymphocytes in mixed lymphocyte culture.

Human CD3- lymphocyte populations were obtained by treating peripheral blood lymphocytes with mAbs directed to CD3, CD4, and CD8 surface antigens. The resulting populations were cultured with irradiated allogeneic cells; at day 4, 100 U/ml IL-2 were added and cultures continued for an additional 10 d. The resulting populations were CD3-CD2+CD7+ and displayed cytolytic activity against PHA-induced blast cells bearing the stimulating alloantigens but not against autologous or unrelated allogeneic blast cells. When CD3- populations were cultured with irradiated autologous cells, no cytolytic activity could be detected either against autologous or allogeneic blast cells. On the other hand, K562 target cells were lysed by both MLC-derived CD3- cell populations regardless of the origin (autologous or allogeneic) of the stimulating cells. CD3- clones were further derived from MLC-stimulated CD3- populations. These clones displayed a cytolytic pattern similar to the original MLC populations as only specific PHA blasts could be lysed. These clones did not express detectable surface TCR-alpha/beta or -gamma/delta molecules and lacked productive mRNA for TCR alpha and beta chains, while small amounts of TCR-gamma mRNA were detectable in one of four clones tested. Also mRNA for CD3 gamma and delta chains were undetectable in all clones, however, CD3 epsilon mRNA was consistently present.

Two subsets of human T lymphocytes expressing gamma/delta antigen receptor are identifiable by monoclonal antibodies directed to two distinct molecular forms of the receptor.

Two mAbs directed to the TCR-gamma/delta were analyzed for their pattern of reactivity with CD3+WT31- cell populations or clones. In normal individuals, the BB3 mAb reacted with approximately 2/3 of peripheral blood CD3+WT31- lymphocytes, whereas delta-TCS-1 stained approximately 1/3 of such cells. In addition, the sum of the percentages of BB3+ and delta-TCS-1+ cells approximated the percentages of peripheral blood CD3+WT31- lymphocytes in seven normal donors tested. Also, in peripheral blood-derived polyclonal CD3+WT31- populations, cultured in IL-2, cells reacting with one or another mAb accounted for the whole cell population. On the other hand, only delta-TCS-1-reactive cells, but not BB3+ cells, could be detected in unfractionated as well as in CD4-8-thymocyte populations. Analysis of peripheral blood-derived CD3+WT31- clones showed that 70% of 72 clones analyzed reacted with BB3 mAb, but not with delta-TCS-1 mAb. On the other hand, delta-TCS-1 mAb stained the remaining BB3- clones. Five clones expressing medium-low amounts of CD8 antigen were BB3- delta-TCS-1+. Both types of clones lysed the Fc gamma receptor-bearing P815 target cell in the presence of anti-CD3 mAb (but not of mAb directed against HLA-DR, CD7 molecules, or TCR-alpha/beta). In this cytolytic assay, BB3 mAb induced target cell lysis only by BB3+ clones, whereas delta-TCS-1 mAb was effective only with delta-TCS-1+ clones. The CD3-associated surface molecules expressed by BB3+ or delta-TCS-1+ clones were analyzed after cell surface iodination and immunoprecipitation with the corresponding anti-TCR mAb or with anti-CD3 mAb (in digitonin-containing buffer). In SDS-PAGE, molecules immunoprecipitated from 13 BB3+ clones displayed, under nonreducing conditions, a molecular weight of 80 kD (in some cases, a minor 38-kD band could be detected). Under reducing conditions, two major components of 44 and 41 kD (and a minor component of 38 kD) were detected. On the other hand, TCR molecules immunoprecipitated from 11 different delta-TCS-1+ clones appeared as a diffuse band of 41-44 kD, both under reducing and nonreducing conditions (under non-reducing condition, an additional 38-kD band was present). Therefore, BB3+ cells express a disulphide-linked form of TCR-gamma/delta whereas delta-TCS-1+ cells express a non-disulphide-linked form.(ABSTRACT TRUNCATED AT 400 WORDS)

Evidence of a natural killer (NK) cell repertoire for (allo) antigen recognition: definition of five distinct NK-determined allospecificities in humans.

Previous studies indicated that CD3-CD16+ natural killer (NK) cells are capable of specific alloantigen recognition. Thus, alloreactive NK clones lysed normal allogeneic target cells (phytohemagglutinin [PHA] blasts) bearing the stimulating alloantigen but did not lyse autologous cells or the majority of unrelated allogeneic cells. In this study we investigated whether NK cells isolated from single individuals could exhibit different allospecificities. To this end, we derived large numbers of CD3-CD16+ clones (in the presence of PHA) from fresh CD3- peripheral blood lymphocytes. Cloning efficiencies ranged between 5 and 10%. The resulting CD3-CD16+ clones were tested for their reactivity against a panel of allogeneic PHA blasts (derived from six donors). In a given individual (A), four distinct groups of clones could be identified according to their pattern of reactivity (over 400 clones have been analyzed). Clones that could be assigned to one or another group of specificity represented 36% of all clones derived from this donor. The remaining clones did not display cytolytic activity against any of the allogeneic target cells used in the panel. None of the clones lysed autologous (A) PHA blasts, yet, these cells were lysed by the representative clones G10 and H12 specific for donor A. Clones displaying a cytolytic pattern of reactivity identical to that defined for donor A were present in other individuals studied, however not all groups of allospecific clones were necessarily represented in different individuals. Allospecific clones belonging to the various groups were homogeneous in the expression of EB6/GL183-triggering surface molecules, and could thus be assigned to one or another of the previously defined subsets of NK cells. Genetic analysis of the new NK-defined alloantigens was performed in representative families. The corresponding characters were found to segregate independently and, at least for three of them, an autosomic recessive type of inheritance could be demonstrated. Moreover, the comparative analysis of the segregation of the major histocompatibility complex haplotypes and the recessive or dominant alleles of the genes governing the five specificities analyzed indicated that there is no independent sampling between the two genetic traits, thus suggesting that the genes regulating the NK-defined specificities are carried by chromosome 6. Finally, some donors expressed more than one specificity, thus providing evidence for an NK-defined complex haplotype.

Human natural killer (NK) alloreactivity and its association with the major histocompatibility complex: ancestral haplotypes encode particular NK-defined haplotypes.

As ancestral haplotypes of the major histocompatibility complex (MHC) appear to define identical MHC haplotypes in unrelated individuals, unrelated individuals sharing the same ancestral haplotype should also share the same NK-defined allospecificities that have recently been shown to map to the human MHC. To test this prediction, multiple cell lines from unrelated individuals sharing the same ancestral haplotypes were tested for the NK-defined allospecificities. It was found that cells sharing the same ancestral haplotypes do have the same NK-defined specificities. Furthermore, the NK-defined phenotype of cells that possess two different ancestral haplotypes can be predicted from the NK-defined phenotypes of unrelated cells that are homozygous for the ancestral haplotypes concerned. Although the group 1 and 2 NK-defined allospecificities can be explained to some extent by HLA-C alleles, evidence is presented that additional genes may modify the phenotype conferred by HLA-C.

Human natural killer cell receptors for HLA-class I molecules. Evidence that the Kp43 (CD94) molecule functions as receptor for HLA-B alleles.

GL183 or EB6 (p58) molecules have been shown to function as receptors for different HLA-C alleles and to deliver an inhibitory signal to natural killer (NK) cells, thus preventing lysis of target cells. In this study, we analyzed a subset of NK cells characterized by a p58-negative surface phenotype. We show that p58-negative clones, although specific for class I molecules do not recognize HLA-C alleles. In addition, by the use of appropriate target cells transfected with different HLA-class I alleles we identified HLA-B7 as the protective element recognized by a fraction of p58-negative clones. In an attempt to identify the receptor molecules expressed by HLA-B7-specific clones, monoclonal antibodies (mAbs) were selected after mice immunization with such clones. Two of these mAbs, termed XA-88 and XA-185, and their F(ab')2 fragments, were found to reconstitute lysis of B7+ target cells by B7-specific NK clones. Both mAbs were shown to be directed against the recently clustered Kp43 molecule (CD94). Thus, mAb-mediated masking of Kp43 molecules interferes with recognition of HLA-B7 and results in target cell lysis. Moreover, in a redirected killing assay, the cross-linking of Kp43 molecules mediated by the XA185 mAb strongly inhibited the cytolytic activity of HLA-B7-specific NK clones, thus mimicking the functional effect of B7 molecules. Taken together, these data strongly suggest that Kp43 molecules function as receptors for HLA-B7 and that this receptor/ligand interaction results in inhibition of the NK-mediated cytolytic activity. Indirect immunofluorescence and FACS analysis of a large number of random NK clones showed that Kp43 molecules (a) were brightly expressed on a subset of p58-negative clones, corresponding to those specific for HLA-B7; (b) displayed a medium/low fluorescence in the p58-negative clones that are not B7-specific as well as in most p58+ NK clones; and (c) were brightly expressed as in the p58+ clone ET34 (GL183-/EB6+, Cw4-specific). Functional analysis revealed that Kp43 functioned as an inhibitory receptor only in NK clones displaying bright fluorescence. These studies also indicate that some NK clones (e.g., the ET34) can coexpress two distinct receptors (p58 and Kp43) for different class I alleles (Cw4 and B7). Finally, we show that Kp43 molecules function as receptors only for some HLA-B alleles and that still undefined receptor(s) must exist for other HLA-B alleles including B27.

Involvement of HLA class I alleles in natural killer (NK) cell-specific functions: expression of HLA-Cw3 confers selective protection from lysis by alloreactive NK clones displaying a defined specificity (specificity 2).

This study was designed to identify the target molecules of the natural killer (NK) cell-mediated recognition of normal allogeneic target cells. As previously shown, the gene(s) governing the first NK-defined allospecificity (specificity 1) were found to be localized in the major histocompatibility complex region between BF gene and HLA-A. In addition, the analysis of a previously described family revealed that a donor (donor 81) was heterozygous for three distinct NK-defined allospecificities (specificities 1, 2, and 5). HLA variants were derived from the B-Epstein-Barr virus cell line of donor 81 by gamma irradiation followed by negative selection using monoclonal antibodies specific for the appropriate HLA allele. Several variants were derived that lacked one or more class I antigen expressions. These variants were analyzed for the susceptibility to lysis by NK clones recognizing different allospecificities. The loss of HLA-A did not modify the phenotype (i.e., "resistance to lysis"). On the other hand, a variant lacking expression of all class I antigens became susceptible to lysis by all alloreactive clones. Variants characterized by the selective loss of class I antigens coded for by the maternal chromosome became susceptible to lysis by anti-2-specific clones. Conversely, variants selectively lacking class I antigens coded for by paternal chromosome became susceptible to lysis by anti-1 and anti-5 clones (but not by anti-2 clones). Since the Cw3 allele was lost in the variant that acquired susceptibility to lysis by anti-2 clones and, in informative families, it was found to cosegregate with the character "resistance to lysis" by anti-2 clones, we analyzed whether Cw3 could represent the element conferring selective resistance to lysis by anti-2 clones. To this end, murine P815 cells transfected with HLA Cw3 (or with other HLA class I genes) were used as target cells in a cytolytic assay in which effector cells were represented by alloreactive NK clones directed against different specificities. Anti-2-specific clones efficiently lysed untransfected or A2-, A3-, and A24-transfected P815 cells, while they failed to lyse Cw3-transfected cells. NK clones recognizing specificities other than specificity 2 lysed untransfected or Cw3-transfected cells. Thus, the loss of Cw3 resulted in the de novo appearance of susceptibility to lysis, and transfection of the HLA-negative P815 cells with Cw3 resulted in resistance to lysis by anti-2 clones. Therefore, we can infer that Cw3 expression on (both human and murine) target cells confers selective protection from lysis mediated by anti-2 NK clones.

Peptide specificity in the recognition of MHC class I by natural killer cell clones.

Recognition by natural killer (NK) cells of major histocompatibility complex (MHC) class I molecules on target cells inhibits NK-mediated lysis. Here, inhibition of NK clones by HLA-B*2705 molecules mutated at single amino acids in the peptide binding site varied among HLA-B*2705-specific NK clones. In addition, a subset of such NK clones was inhibited by only one of several self peptides loaded onto HLA-B*2705 molecules expressed in peptide transporter-deficient cells, showing that recognition was peptide-specific. These data demonstrate that specific self peptides, complexed with MHC class I, provide protection from NK-mediated lysis.

CD1d expression on B-precursor acute lymphoblastic leukemia subsets with poor prognosis.

Acute lymphoblastic leukemia (ALL) is the most frequent malignancy of childhood. Although therapeutical advances have been achieved, some ALL subgroups still fare poorly. CD1d is a monomorphic molecule that provides a suitable target for immunotherapy in view of the characterization of a glycolipid, alpha-galactosylceramide (alpha-GalCer), capable of being presented to CD1d-restricted T cells with cytotoxic potential. We investigated CD1d expression in 80 pediatric B-cell precursor (BCP) ALL cases defined according to immunophenotype, cytogenetic features and age at onset. CD1d was detected on ALL cells in 15% of the patients. CD1d+ ALLs were significantly associated with infant leukemia, pro-B phenotype and mixed-lineage leukemia (MLL)/AF4 gene rearrangement. Accordingly, overall survival of patients with CD1d+ ALL was significantly shorter. CD1d+ leukemic blasts were able to present alpha-GalCer via CD1d to cytotoxic CD1d-restricted T cells, which induced apoptosis of ALL cells that was inhibited by mAb to CD1d. CD1d+ blasts loaded with alpha-GalCer elicited cytokine secretion by CD1d-restricted T cells. Analysis of bone marrow (BM) cells derived from normal donors revealed that CD19+/CD1d+ cells were mostly mature B lymphocytes. However, a minority of BCPs expressed CD1d. Thus, expression of CD1d in ALL cases heralds an adverse prognosis but may provide a therapeutic tool.